colpitt osc at UHF

[PDRUNEN]

Hi Group,

Designing a 372MHz colpitt but just can not get the FCC allowed max power at
372 without the 3 harmonic (inside the restricted band) going over the limit.

Currently I have used the UHF NPN transistors but would ask if I could get
better preformance using a UHF FET?

I currently use 3V on the circuit but would the harmonics be helped if the
power supply was +/- 1.5V or +/- 3.0V?

How about a different configuration? I use the common emitter but would a
common collector or base configuration do better?

The colpitt is such a small circuit, a saw resonator, a couple of caps and
resistors, a circuit this simple is so hard to tame!!!

Thanks!

Paul

 

[Tim Wescott]

PDRUNEN wrote:

Hi Group,

Designing a 372MHz colpitt but just can not get the FCC allowed max power at
372 without the 3 harmonic (inside the restricted band) going over the limit.

Currently I have used the UHF NPN transistors but would ask if I could get
better preformance using a UHF FET?

I currently use 3V on the circuit but would the harmonics be helped if the
power supply was +/- 1.5V or +/- 3.0V?

How about a different configuration? I use the common emitter but would a
common collector or base configuration do better?

The colpitt is such a small circuit, a saw resonator, a couple of caps and
resistors, a circuit this simple is so hard to tame!!!

Thanks!

Paul

Trying different configurations may work. Have you tried a low-pass
filter or a wavetrap?

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Jim Thompson]

On 29 Oct 2004 23:21:33 GMT, pdrunen@aol.com (PDRUNEN) wrote:

Hi Group,

Designing a 372MHz colpitt but just can not get the FCC allowed max power at
372 without the 3 harmonic (inside the restricted band) going over the limit.

Currently I have used the UHF NPN transistors but would ask if I could get
better preformance using a UHF FET?

I currently use 3V on the circuit but would the harmonics be helped if the
power supply was +/- 1.5V or +/- 3.0V?

How about a different configuration? I use the common emitter but would a
common collector or base configuration do better?

The colpitt is such a small circuit, a saw resonator, a couple of caps and
resistors, a circuit this simple is so hard to tame!!!

Thanks!

Paul

An MC12148 (one of my ancient designs will give you a high purity
oscillator, but probably won't give the power you need without a
buffer.

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice480)460-2350 | |
| E-mail Address at Website Fax480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

 

[John S. Dyson]

In article <hg06o0dhidc88teadp2fi56tumd1maahq4@4ax.com>,
John Larkin <jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> writes:

On 29 Oct 2004 23:21:33 GMT, pdrunen@aol.com (PDRUNEN) wrote:

Hi Group,

Designing a 372MHz colpitt but just can not get the FCC allowed max power at
372 without the 3 harmonic (inside the restricted band) going over the limit.

Currently I have used the UHF NPN transistors but would ask if I could get
better preformance using a UHF FET?

I currently use 3V on the circuit but would the harmonics be helped if the
power supply was +/- 1.5V or +/- 3.0V?

How about a different configuration? I use the common emitter but would a
common collector or base configuration do better?

The colpitt is such a small circuit, a saw resonator, a couple of caps and
resistors, a circuit this simple is so hard to tame!!!

Thanks!

Paul

I've done Colpitts at 600 MHz, using an HJ fet, and wasn't very happy
with the performance. I'm not an oscillator expert, but I'm thinking
that the tight coupling of the low-impedance source (or emitter) into
the guts of the resonator messes up the Q pretty badly. Randy Rhea's
book on RF oscillators is very good, and he has a lot of examples with
transistor or MMIC gain stages lightly coupled into/out of the
resonator node, usually with small capacitors.

I'd consider using BJT technology MMIC, with SiGe as first choice

because of better 1/F noise. GaAs would be workable, with perhaps
more 1/F noise, so I'd tend to prefer a selected Si part instead
of GaAs.

Some of the GaAs and SiGe MMIC amps are dirt cheap, have lots of GBW,
and are very repeatable. They can make a nice oscillator.

At 600MHz, you really don't need LOTS of GBW, that is almost DC for

GaAs, SiGe type technology.

HJ Fets would certainly oscillate, but might be a little tricky to
control (but one also has to be careful with any fast technology), but
HJ Fets also tend to have more noise than SiGe.

If possible, you might really prefer to stick with Si only, especially
at 600MHz. For an oscillator, I most likely wouldn't use a component
with a 45GHz fT, unless it is SiGe. The tradeoff of trying to tame
the component might be worth the lower noise.

John

 

[Paul Burridge]

On Fri, 29 Oct 2004 19:58:17 -0700, John Larkin
<jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

Some of the GaAs and SiGe MMIC amps are dirt cheap, have lots of GBW,
and are very repeatable. They can make a nice oscillator.

Got any part numbers, John?
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[John Larkin]

On Sat, 30 Oct 2004 12:13:54 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

On Fri, 29 Oct 2004 19:58:17 -0700, John Larkin
jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

Some of the GaAs and SiGe MMIC amps are dirt cheap, have lots of GBW,
and are very repeatable. They can make a nice oscillator.

Got any part numbers, John?

MiniCircuits ERA-series InGaAs parts (they have nice sample kits) and
the Sirenza SGA-series SiGe amps. The Sirenzas are very quiet. Tons of
other people make similar stuff, up to maybe 10 GHz. These make nice
wideband (pulse/time domain) amps, too.

John

 

[Paul Burridge]

On Sat, 30 Oct 2004 08:44:46 -0700, John Larkin
<jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

MiniCircuits ERA-series InGaAs parts (they have nice sample kits) and
the Sirenza SGA-series SiGe amps. The Sirenzas are very quiet. Tons of
other people make similar stuff, up to maybe 10 GHz. These make nice
wideband (pulse/time domain) amps, too.

Thanks, John. It's obviously about time I got better acquaited with
MiniCircuits and their stuff. I shall so do!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[dd]

In message <g6o7o0pb97rv0nq19r6tveuths85mb9i7m@4ax.com>, Paul Burridge
<pb@notthisbit.osiris1.co.uk> writes

On Sat, 30 Oct 2004 08:44:46 -0700, John Larkin
jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

MiniCircuits ERA-series InGaAs parts (they have nice sample kits) and
the Sirenza SGA-series SiGe amps. The Sirenzas are very quiet. Tons of
other people make similar stuff, up to maybe 10 GHz. These make nice
wideband (pulse/time domain) amps, too.

Thanks, John. It's obviously about time I got better acquaited with
MiniCircuits and their stuff. I shall so do!
No-one has defined the real reason for your problem

The oscillation is building up to the level where the waveform is
clipped.
You need this excess gain to start oscillation but once oscillation is
started it is necessary to back off oscillator gain by means of a
control loop to the point that the fet/bjt is in its linear region.

--
dd